9.

4 Frontal advance for unsteady 1D displacement The unsteady state displacement of oil by water is due to the change in water saturation with time. This can be visualized by looking at the schematics in Figure 9.6. These schematics represent

Sor Sw Swi Sor Sw Swi 0 x/L

Sor Sw Swi

Sor Sw Swi 1 0 x/L

Figure 9.6 Progression of water displacing oil for immiscible, 1D snapshots in time of the frontal boundary as water is displacing oil. In sequence, A depicts the initial state of the sample (or reservoir) where saturations are separated into irreducible water, residual oil and mobile oil components. After a given time of injection, the front advances to a position as shown in B. Ahead of the front water saturation is at irreducible, but behind the front water saturation is increased. Continuing in time, eventually the water will breakthrough the end of the core (reservoir) and both oil and water will be produced simultaneously, C. Continued injection will increase the displacing phase saturation in the core (reservoir), D. Two methods to predict the displacement performance are 1) the analytical solution by Buckley Leverett (1941), and 2) applying numerical simulation. Only the analytical solution will be described in this chapter. 9.4.1 Buckley Leverett (1941) The derivation begins from the 1D, multiphase continuity equations.

 (ouox ) = (o So ) x t

(9.19) (9.20)

( wuwx ) = ( w S w ) t x

In terms of volumetric flow rate,

(o qo ) = A (o So ) t x ( wqw ) = A ( w S w ) t x

(9.21) (9.22)

Assume the fluids are incompressible and the porosity is constant. Eqs. (9.21) and (9.22) simplify to,
qo S = A o x t qw S = A w x t

(9.23) (9.24)

Combining,
(S w + S o ) (qw + qo ) = A =0 t x

(9.25)

The result is qT = qo + qw = constant. That is, the total flow rate is constant at each crosssection. From the definition of fractional flow, qw = f w qT qo = (1 f w )qT Substitute into Darcys equation for each phase, (9.26)

qo = (1 f w )qT =
qw = f w qT =

ko A po + o g sin o x

(9.27) (9.28)

k w A pw + w g sin w x dp o dx dp w dx

Rearranging Eqs. (9.27) and (9.28), we can substitute for the pressure gradient terms.

dp c dx

Solving the resulting equation provides the complete fractional flow equation.

ko A pc + g sin 1 q x fw = + o T k k 1+ o w 1+ o w k w o k w o

(9.28)

In the analytical solution it is difficult to analyze the derivative term (dpc/dx). If we expand this derivative to,

pc S w pc = x S w x
and assume dpc/dx 0, then either term on the right-hand side must be small and

(9.29)

approach zero as well. In linear displacement, dpc/dSw 0 at moderate to high water saturations. This is evident by inspection of a capillary pressure curve such as shown in Figure 9.7.
pc 0 S w

Pc

Sw

Figure 9.7 Capillary pressure curve illustrating flat transition region at moderate to high water saturations. If the derivative term is negligible, and flow is in the horizontal direction such that no gravity term is present, then the fractional flow equation reduces to,
fw = 1 k 1+ o w k w o
k w o ko w

(9.30)

If we define mobility ratio as,

M =

(9.31)

then fw = 1/(1+1/M).

If we return to Eq. (9.24) and substitute for qw, we obtain,

f w A S w = x qT t

(9.32)

To develop a solution, Eq. (9.32) must be reduced to one dependent variable, either Sw or fw. Observe, Sw = Sw(x,t) or, S w dS w = x S dx + w t t dt x (9.33)

By tracing a fixed saturation plane through a core, that is,

dS w ( x, t ) dt Sw

= 0 , then

dx dt S w

S w t x = S w x t

(9.34)

where the right-hand side is the velocity of the saturation front as it moves through the porous media. Observe fw = fw(Sw) only, then,
f w S w f w = x t S w t x t

(9.35)

Substitution of Eqs. (9.34) and (9.35) into Eq. (9.32), results in the frontal advance
equation.

dx dt

=
Sw

qT f w A S w t

(9.36)

Equation (9.36) represents the velocity of the saturation front. Basic assumptions in the derivation are incompressible fluid, fw(Sw) only and immiscible fluids. Furthermore, only oil is displaced; i.e., the initial water saturation is immobile, and no initial free gas saturation exists; i.e., not a depleted reservoir. The location of the front can be determined by integrating the frontal advance equation,
xS w

dx S =
0
w

t 1 f w S qT dt A w t 0

(9.37)

If injection rate is constant and if the dfw/dSw = f(Sw) only, then

xS =
w

qT t f w A S w S

(9.38)
w

We can evaluate the derivative from the fractional flow equation (Eq. 9.28), either graphically or analytical. Figure 9.8 illustrates the graphical solution.

Swbt fwf fw

Swc Sw

Swf

Figure 9.8 Fractional flow curve The fractional flow of water at the front, fwf, is determined from the tangent line originating at Swc. The corresponding water saturation at the front is Swf. The average water saturation behind the front at breakthrough, Swbt, is given by the intersection at fw = 1. The location of the front is determined by Eq. (9.38), with the slope of the tangent to the fractional curve used for the derivative function. 9.4.2 Displacement Performance (constant injection rate) The displacement performance can be separated into two intervals, before and after breakthrough. Until breakthrough, the volume of oil produced is equal to the volume of water injected. After breakthrough, water saturation gradients exists, thus the volume of water in the system can be defined as;
Vw = S w Adx
x1 x2

(9.39)

and the volume of oil displaced,

Vo = Vw A ( x2 x1 ) S wi

(9.40)

Figure 9.9 illustrates the recovery of oil both before and after water breakthrough. Note the linear slope up to breakthrough and then the decrease in slope (reduced performance) after breakthrough.

Np

breakthrough

Qi

Figure 9.9 Typical oil recovery performance plot for immiscible displacement A solution for waterflood performance was developed by Welge in 1952. Define the volumetric average water saturation as,
x2

Sw =

x1 x2

S
x1

Adx

(9.41)

Adx

For constant cross-sectional area (A) and porosity (), Equation (9.41) reduces to,
x2

Sw =

x1

dx

x2 x1

(9.42)

The integrand can be expanded and the equation rearranged such that,
x S xS 1 S w = 2 w 2 1 w1 x dS w x2 x1 x2 x1 1
2

(9.43)

Substitute the frontal advance equation (Eq. 9.35) for the integral and solve,
2 qT t f w x dSw = dS w A 1 1 S w S 2
w

qT t ( f w2 f w1 ) A

(9.44)

Thus the general Welge equation is,

Sw =

x2 S w 2 x1S w1 qT t f w 2 f w1 x2 x1 A x2 x1

(9.45)

A useful simplification is to consider x1 = 0 at the inlet and x2 = L at the outlet end of the core,
Sw = Sw2 + qT t (1 f w2 ) AL

(9.46)

where fw1 is assumed to be one at the inlet. Define the total volume injected, Wi, = qT*t, and the pore volume, Vp = AL. Combining gives the number of pore volumes injected, Qi,
Qi = Wi Vp

(9.47)

Thus we can write Eq. (9.46) in terms of Qi.

S w = S w 2 + Qi (1 f w 2 )

(9.48)

The cumulative oil displaced, Np, can be expressed in terms of the difference in the average water saturation and the exit end saturation, i.e.,

N p = Vp Sw Sw2

(9.49)

Consider a special case immediately before breakthrough. In this case, Sw2 = Swi and fw2 = 0. Subsequently, Eq. (9.48) can be written as:
S wbt S wi = Qibt

(9.50)

and the cumulative oil displaced:

N p = V p S wbt S wi

(9.51)